[nostrademons]

I think your conclusion is probably right on. You can do everything right and still fail in the startup world, because there's a large element of luck. In Sam/Loopt's case, I'm almost certain they were early - FourSquare/Latitude/Facebook-Check-Ins did okay just 4 years later. But then, didn't someone once say "Early is wrong as far as startups go." Perhaps that means you shouldn't listen to sama's advice on market-timing, but it doesn't mean you shouldn't listen to his advice on the other aspects of startups.

The more heartening way of looking at your conclusion is that if you do everything right - don't screw anyone over, take calculated prudent risks, seek out information, make sane decisions, and work hard - and if you make that visible to people, you can usually soft-land even if your startup fails. That was my experience with my own failed startup - once I folded it up and wrote a postmortem, I had a large number of job offers, and took one at Google. People judge you based on what you do, not for things out of your control.